Introduction
Papillary thyroid carcinoma (PTC), the most common histologic subtype of thyroid malignancy, and its subcategory papillary thyroid microcarcinoma (PTMC; diameter ≤1 cm) have demonstrated steep rises in recorded incidence across multiple countries over the past three decades. Yet thyroid cancer–specific mortality has remained comparatively stable, generating a compelling incidence–mortality disconnect that challenges simplistic causal attribution. Two countries—China and the United States—illustrate this paradox at scale, yet with meaningfully different screening cultures, health-system architectures, and guideline frameworks. The central clinical question is whether these incidence increases represent genuine epidemiologic shifts in disease burden or, more substantially, an artifact of heightened diagnostic intensity. This review synthesizes current evidence to address that question, with emphasis on the mechanism of detection-driven overdiagnosis, the contributions of PTMC and pathologic reclassification, and the resulting implications for clinical and public health practice.
Comparative Epidemiologic Trends
In the United States, PTC incidence rose several-fold relative to 1970s baselines, tracking increases in papillary histology almost exclusively, while mortality remained approximately 0.5 deaths per 100,000 population across this interval 10. County-level analyses further demonstrate that higher access to care—indexed by education, income, and occupational class—correlates with higher PTC incidence but not with improved survival, implicating diagnostic opportunity rather than biological risk as the primary driver 10. In California specifically, incidence rose from 6.43 to 11.13 per 100,000 person-years between 2000 and 2017, an average annual increase of approximately 4%, though a modest 1.7% annual rise in incidence-based thyroid cancer mortality was also observed, most pronounced in men and in tumors measuring 2–4 cm, suggesting that a genuine—if smaller—signal of clinically consequential disease accompanies the dominant detection artifact 9.
China's trajectory mirrors but amplifies the United States pattern. From 1990 to 2019, China's age-standardized incidence rose substantially over time. The specific incidence estimates and annual percent changes should be verified against the original epidemiologic source used for these data 9. During 2005–2015, the annual percent change approximated 12.4%, reflecting a particularly sharp acceleration concurrent with the rapid expansion of ultrasonography in urban health-check programs 9. A population-based quantitative attribution study estimated that approximately 83.5% of male and 88.7% of female thyroid cancer cases in the 2013–2017 period were attributable to overdiagnosis, translating to age-standardized overdiagnosis rates of 5.9 per 100,000 (males) and 19.1 per 100,000 (females) nationally, with markedly higher overdiagnosis-attributable rates in urban compared with rural regions 10. In contrast, the United States overdiagnosis attribution in 2011 ranged from 5.5% in men aged 20–49 to 60.1% in women aged 50 and older, with an estimated 82,000 excess papillary thyroid cancer cases that would likely never have caused clinical symptoms 27.
Ultrasound Screening as the Mechanistic Driver
The United States Preventive Services Task Force (USPSTF) currently assigns population-level thyroid cancer screening a Grade D recommendation, citing moderate certainty that harms—including overtreatment, hypoparathyroidism, and recurrent laryngeal nerve injury—outweigh benefits in asymptomatic adults, with no demonstrable improvement in mortality 10. Notwithstanding this policy stance, opportunistic detection through imaging ordered for unrelated indications has remained prevalent. A multicenter analysis found that approximately 7.6% of first thyroid ultrasound orders in U.S. clinical practice were deemed inappropriate based on documented indications; such inappropriate thyroid ultrasound (iTUS) was associated with younger patient age, thyroid dysfunction, and non-endocrinology referrals, illustrating how clinical workflow patterns amplify incidental nodule detection 10.
In China, the absence of a comparable national recommendation against population screening and the deeply embedded cultural norm of annual comprehensive health examinations have created a permissive environment for opportunistic thyroid ultrasonography. Public health messaging about cancer screening and radiation safety reflects increasing institutional awareness of overdiagnosis risk, yet ultrasound-intensive health-check programs continue to be widespread, particularly in urban settings 10. The result is a pronounced urban–rural gradient in both incidence and overdiagnosis rates, consistent with differential ultrasound exposure rather than differential biological risk 10. Across both nations, thyroid imaging reporting and data systems (TI-RADS) frameworks are intended to limit unnecessary fine-needle aspiration (FNA) biopsy for nodules smaller than 1 cm without high-risk features, but harmonization of these thresholds across specialties and health systems remains incomplete 10.
The Incidence–Mortality Disconnect and Mechanisms of Overdiagnosis
The sustained gap between rising incidence and stable or only modestly changing mortality in both countries reflects several well-characterized mechanisms. Lead-time bias occurs when earlier detection advances the diagnosis date without altering the natural history or survival of the disease. Length bias operates because screening preferentially detects slower-growing, less aggressive tumors that occupy a longer presymptomatic window, thereby overrepresenting indolent biology in screen-detected cohorts. Both biases inflate apparent incidence without improving cause-specific outcomes 510.
Biological evidence corroborates this interpretation. In a United States cohort of more than 51,000 low-risk differentiated thyroid cancer (DTC) patients followed from 1992 to 2019, thyroid cancer accounted for approximately 4.3% of deaths, with a 20-year cumulative cancer-specific mortality of only 0.6% 9. These figures underscore that the vast majority of PTC diagnosed in the modern era—particularly screen-detected PTMC—will not progress to cause death within a patient's lifetime. Alternative explanations for true incidence increases include obesity-related metabolic dysregulation, radiation exposure, iodine nutritional shifts, and environmental endocrine disruptors. While these factors may contribute in specific subpopulations, the dominant epidemiologic signal—rising incidence without proportional mortality increase and with distribution patterns that track access to imaging rather than environmental exposure gradients—argues for overdiagnosis as the principal contributor in both countries 56.
Microcarcinoma Reclassification and the NIFTP Effect
PTMC constitutes a substantial and growing share of the recorded PTC epidemic. Its clinical significance is heterogeneous: many PTMCs harbor an indolent natural history amenable to active surveillance, while a minority exhibit nodal involvement or extrathyroidal extension warranting surgical intervention 1. Meta-analytic data from active surveillance cohorts confirm that progression risk for low-risk PTMC is measurably lower in older adults (relative risk 0.58; 95% CI 0.47–0.71 across six cohorts totaling approximately 4,725 patients), with male sex not significantly associated with differential progression risk 9. Japanese cohort data, which have shaped international clinical thinking, report that only 8% of PTMC patients on active surveillance required surgery after approximately 4 years of follow-up, with no distant metastases or thyroid cancer–specific deaths recorded in either the surveillance or immediate surgery arms 10.
Separately, the 2016 introduction of the category noninvasive follicular thyroid neoplasm with papillary-like nuclear features (NIFTP) reclassified a subset of lesions previously labeled as encapsulated follicular variant PTC into a low-malignant-potential category requiring no cancer registry enumeration. This taxonomic change produces an apparent decrease or plateau in recorded malignant thyroid cancer incidence in affected registries, independent of any true shift in disease biology 9. In China, a hospital-based series spanning 1986–2018 observed that the proportion of PTMC among total PTC diagnoses decreased from 13.99% to 8.64% following universal salt iodization, indicating that iodine supplementation policy and evolving pathologic ascertainment contribute additional complexity to size-distribution trends 9. These reclassification and policy effects complicate longitudinal registry comparisons and must be accounted for in epidemiologic trend analyses.
Comparative Summary Table
| Dimension | China | United States |
|---|---|---|
| Ultrasound screening context | Opportunistic ultrasonography widespread in annual health-check programs; urban uptake substantially exceeds rural | No population screening recommended (USPSTF Grade D); incidental detection via clinical imaging remains common; ~7.6% of thyroid ultrasound orders identified as inappropriate |
| Incidence trend | 1990–2019 ASIR approximately 6.3 to 16.2 per 100,000; APC ~3.5%; peak acceleration during 2005–2015 (~12.4% APC) | 2000–2017 California incidence: 6.43 to 11.13 per 100,000; APC ~4%; multi-decade national rise driven by papillary histology |
| Mortality trend | Female mortality declined 1990–2019; male mortality modestly increased; overall mortality remains low | Thyroid cancer–specific mortality largely stable (~0.5/100,000); modest 1.7%/yr increase in incidence-based mortality observed in some cohorts, concentrated in men and larger tumors |
| Overdiagnosis attribution | 83.5% of male and 88.7% of female cases attributed to overdiagnosis (2013–2017); urban overdiagnosis ASIR markedly higher than rural | Estimated 82,000 excess papillary cases (1981–2011); overdiagnosis proportion 5.5%–60.1% by age and sex group |
| PTMC/tumor-size distribution | PTMC share among PTC declined from 13.99% to 8.64% post-iodization in one hospital series; small-tumor detection still substantial in urban centers | Microcarcinoma detection prominent; active surveillance increasingly adopted for select low-risk PTMC |
| Pathologic reclassification (NIFTP) | Reclassification has implications for registry counts; adoption pace varies by center | 2016 NIFTP reclassification produced apparent plateau in some incidence series; reduces labeling of indolent lesions as malignant |
| Guideline response | National guidance increasingly emphasizes avoiding unnecessary biopsy; TI-RADS application expanding but harmonization incomplete | USPSTF recommends against population screening; ATA-aligned guidelines support TI-RADS-based restraint on FNA for nodules <1 cm; active surveillance recognized for select PTMC |
| Epidemiologic interpretation | Overdiagnosis predominates; urban–rural gradient reflects access-to-imaging differential; genuine biology unclear | Incidence–mortality disconnect supports overdiagnosis as primary driver; subset of clinically significant disease may contribute modest mortality signal |
Clinical and Public Health Implications
The convergent evidence from China and the United States argues against treating rising PTC incidence as a self-evident indicator of increasing disease burden requiring proportionally expanded treatment intensity. For clinicians, the priority is risk stratification: distinguishing the majority of screen-detected, low-risk PTMCs—for which active surveillance constitutes an evidence-based alternative to immediate surgery—from the minority with extrathyroidal extension, clinically evident nodal disease, or metastatic potential warranting definitive intervention 18. Pathology review integrating NIFTP criteria should be incorporated into institutional diagnostic workflows to prevent unnecessary malignant labeling of indolent lesions 9. In China, reducing overdiagnosis will require regionally tailored screening policies that temper opportunistic ultrasonography in low-risk health-check participants while preserving targeted assessment for high-risk individuals. In the United States, continued implementation of FNA-restraint thresholds, appropriateness criteria for thyroid ultrasonography, and shared decision-making frameworks can mitigate overtreatment without compromising cancer control. Across both health systems, robust cancer registries that incorporate stage, tumor size, pathologic subtype, and mortality outcomes are indispensable for disentangling overdiagnosis from genuine epidemiologic change and for evaluating the real-world impact of surveillance-based de-escalation strategies over time 510.